{"id":"W2906501467","doi":"10.15353/jcvis.v4i1.329","title":"On Robustness of Deep Neural Networks: A Comprehensive Study on the Effect of Architecture and Weight Initialization to Susceptibility and Transferability of Adversarial Attacks","year":2018,"lang":"en","type":"article","venue":"Journal of Computational Vision and Imaging Systems","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Canadian Institute for Advanced Research","keywords":"Robustness (evolution); Initialization; Transferability; Computer science; Adversarial system; Artificial neural network; Network architecture; Artificial intelligence; Network model; Architecture; Machine learning; Data mining; Computer security","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006509411,0.002232153,0.001122497,0.001639439,0.0006206565,0.001293007,0.001079873,0.001546043,0.001831779],"category_scores_gemma":[0.04147127,0.0006576041,0.001192664,0.0008387279,0.001801676,0.003380871,0.002170696,0.002656735,0.0003127106],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001437786,"about_ca_system_score_gemma":0.0007594603,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002243373,"about_ca_topic_score_gemma":0.001822346,"domain_scores_codex":[0.9964574,0.001344149,0.0003381252,0.000525247,0.0009764778,0.0003586794],"domain_scores_gemma":[0.9580013,0.03199278,0.003527369,0.004369487,0.00165582,0.0004532088],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000389906,0.0001512364,0.003817236,0.00034893,0.000334323,0.0001428484,0.00007870051,0.9287296,0.009087333,0.004518242,0.0006583207,0.05174325],"study_design_scores_gemma":[0.00001952888,0.0007374762,0.003147132,0.0001673403,0.0001652667,0.0002241528,0.00006660091,0.9573901,0.02753893,0.009426382,0.0010627,0.00005436896],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5519207,0.01383375,0.4195271,0.001851464,0.0003503016,0.0003096478,0.0005405101,0.001706955,0.009959643],"genre_scores_gemma":[0.9796751,0.002613639,0.01572123,0.0001517984,0.00007179063,0.00006602942,0.0002539215,0.0001620211,0.001284458],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006509411,"threshold_uncertainty_score":0.03442544,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009030126548489809,"score_gpt":0.2898208874482386,"score_spread":0.2807907608997487,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}